Hand-Eye Calibration

This package is used for hand-eye calibration. Intel Realsense D455 camera and JAKA MiniCobo manipulator are required.

Usage

Step 1: Print a calibration target

This package supports a variety of calibration targets, including checkerboard, single ArUco marker, ArUco bundles, single AprilTag marker, AprilTag bundles and ChArUco marker. You can find different types of markers through the links below. You can also use our provided Python scripts to generate custom markers.

• Single ArUco marker
• ArUco bundles
• Single AprilTag marker

Step 2: Modify the documents in launch and config

Step 3: Enable the camera

roslaunch realsense2_camera rs_camera.launch

Step 4: Marker detection and pose estimation

• Checkerboard

roslaunch hand_eye_calibration checkerboard_start.launch

• Single ArUco marker

roslaunch hand_eye_calibration aruco_single_start.launch

• ArUco bundles

roslaunch hand_eye_calibration aruco_bundle_start.launch

• Single AprilTag marker

roslaunch hand_eye_calibration apriltag_single_start.launch

• AprilTag bundles

roslaunch hand_eye_calibration apriltag_bundle_start.launch

• ChArUco

roslaunch hand_eye_calibration charuco_start.launch

Step 5: Enable the manipulator

roslaunch hand_eye_calibration robot_start.launch

Step 6: Hand-eye calibration.

There are two options for hand-eye calibration: manual and automatic (without human assistance). If you choose the former one, please move the manipulator with your hands during calibration. If you choose the latter one, please predefine the trajectories of the manipulator before calibration.

• Manual

rosrun hand_eye_calibration hand_to_eye_calib.py

• Automatic

rosrun hand_eye_calibration automatic_hand_to_eye_calib.py